Funny story of a trivial repair attempt over an electronic ballast used to drive a fluorescent UV bulb for an insect trap.
DISCLAIMER !!!
This stuff is connected DIRECTLY to the power grid and the fluorescent bulb emits dangerous UV !!
Working on these type of circuit REQUIRES knowledge and precautions !
DON'T ATTEMPT ANY REPAIR UNLESS YOU KNOW WHAT YOU ARE DOING !
The trap consist in a UV fluorescent bulb that is used to attract insects.
On the back of the lamp, there is an holder for cartridges that capture (glue) the insects.
As stated before, the UV emitted by the bulb CAN DAMAGE SKIN AND EYES.
Usually the bulb is not visible directly when the trap is put together, however in order to repair it, the bulb can be fully exposed and thus UV can reach you !
The problem
The lamp remains OFF when the trap is plugged in.
After opening the trap (5 screws to remove the rear cover) the circuit, an electronic ballast, is exposed.
A quick analysis shows the problem :
The two resistors are connected to the base of two MJE13003, a fast switching power transistor, very probably destroyed as well.
The fuse is still intact and doesn't need to be replaced.
With that and the help of other people on the net I did find a schematic that is SIMILAR to the circuit to repair :
The used transistor in the schematic is the MJE13001 but it should be relatively safe to assume the same base resistor value, here placed to 20 Ohm 1/4 Watt (R5 and R2 on the schematic).
Now let see the PCB component side after the removal
In the last photo the damage around Q1 is clearly visible even after the clean up, sign that Q1/R4 generated a LOT of heat for quite long time.
Very possible on design .. so that one in the end is forced to buy a new unit.
A repair would imply to :
- determine the original value of the resistors (impossible to read anything on a so damaged components)
- replace the burned resistors
- replace the two transistors MJE13003, as said they are very probably destroyed anyway
If anything else is not damaged, the circuit should starts to work again.
Diagnosis
There are 3 major obstacles to the repair :
- know the resistors value
- hoping the PCB will survive the component removal since the heat could make the traces to detach and the Bakelite to become brittle
- find the MJE13003 (is obsolete - not in production anymore)
Resistor value
As suggested in a forum, sometime one can be lucky and simply cutting out one terminal of the resistor in order to lift it from the PCB, is possible to read the value below.
Cutting one wire rather than desolder it, can prevent the resistor to fall apart if it is too much burnt out.
Sometime the burning can happens on the top but on the bottom some colored bands could be still readable.
Cutting one wire rather than desolder it, can prevent the resistor to fall apart if it is too much burnt out.
Sometime the burning can happens on the top but on the bottom some colored bands could be still readable.
Of course I'm not so lucky, both resistors are well burned also on the bottom, on contact with the PCB.
 |
| One of the two resistors has at least a red band visible, so could be something starting with 2 |
 |
| This one is TOTALLY burned out |
The first attempt to find the value thus miserably failed.
In order to figure out the value for the base resistor I started to a quick reverse engineering of the circuit (just the transistor configuration) and then look around on the net for similar configurations.
With that and the help of other people on the net I did find a schematic that is SIMILAR to the circuit to repair :
The used transistor in the schematic is the MJE13001 but it should be relatively safe to assume the same base resistor value, here placed to 20 Ohm 1/4 Watt (R5 and R2 on the schematic).
Plus another guy sent to me an interesting tutorial on electronic ballast, about 380 pages in .. Russian !
With some patience I was able to translate enough to compare with some schematics, this time based on the MJE13003.
With some patience I was able to translate enough to compare with some schematics, this time based on the MJE13003.
These schematics did show a value of 22 Ohm for the base resistors and 1 Ohm for the emitter resistor, like the ones on the circuit to repair.
PCB
Time to remove the burned components and asses the PCB health.
 |
| The PCB in the PCB holder |
After putting the PCB in a PCB holder I was able to remove the transistors and the burned resistors easily, so far no damage to the traces.
 |
| The damaged components. I would say the transistors are gone for good as well |
 |
| The traces are in a very good shape, a little bit change of color of the traces around the burned area, but they are not detaching from the Bakelite. |
Now let see the PCB component side after the removal
 |
| Lot of smoke dust, lets clean it up and see the PCB conditions |
 |
| Not bad. Q1 is where the damage is bigger, bad burn but didn't go through, the traces below are still in good conditions. |
In the last photo the damage around Q1 is clearly visible even after the clean up, sign that Q1/R4 generated a LOT of heat for quite long time.
Very possible on design .. so that one in the end is forced to buy a new unit.
Checked visually also all the other components for hidden damage, especially the ones around the burned ones, but everything seems in pristine conditions.
For now I'll assume no other components are damaged.
For now I'll assume no other components are damaged.
The repair
Time to attempt the repair.
As said before, a 20 Ohm resistor (or around such value) should do the trick.
Since I have around some 22 Ohm 1/4 Watt resistors I'll use a couple of them, then the two MJE13003 replacement.
I'm assuming that for such cheap circuits a 5% tolerance resistors are used, so 20/22 in the end is the same.
I did order some MJE13003 from China because this transistor is out of production for quite some time now (2013) and is not easy to found it at low cost elsewhere.
It doesn't make sense to spend more than 5$ each as I found on eBay for example !!!
I ordered 10 from China for about 4$, shipment included.
The usual suppliers (DigiKey, Mouser) doesn't have it anymore.
This means to wait some time before to receive them.
However looking in one of my junk boxes, I did find two ST13003, basically identical to the MJE13003 but from SGS.
The first attempt then will be using a couple of ST13003 and see if they survive.
Of course I'll use a new bulb as well since it is VERY probable the burnout happened for a discharge/short on the old bulb (burned out too).
I'm assuming that for such cheap circuits a 5% tolerance resistors are used, so 20/22 in the end is the same.
I did order some MJE13003 from China because this transistor is out of production for quite some time now (2013) and is not easy to found it at low cost elsewhere.
It doesn't make sense to spend more than 5$ each as I found on eBay for example !!!
I ordered 10 from China for about 4$, shipment included.
The usual suppliers (DigiKey, Mouser) doesn't have it anymore.
This means to wait some time before to receive them.
However looking in one of my junk boxes, I did find two ST13003, basically identical to the MJE13003 but from SGS.
The first attempt then will be using a couple of ST13003 and see if they survive.
Of course I'll use a new bulb as well since it is VERY probable the burnout happened for a discharge/short on the old bulb (burned out too).
Round 1
The first score is for the circuit, after replacing resistors and the transistors, is still not working.
So something else is broken down.
Capacitors, diode and coils and other resistors seems in perfect orders, where possible I did check them.
Capacitors, diode and coils and other resistors seems in perfect orders, where possible I did check them.
The only component not tested/checked is the DIAC.
A failed DIAC of course would prevent the oscillator to start obtaining exactly the result obtained.
Thus the next step is to fix the DIAC with a new one.
Thus the next step is to fix the DIAC with a new one.
Of course this means more waiting to receive them since I don't have any DIAC around 👿
Round 2
... and here we are !!!
Also changing the DIAC nothing is working.
So this time the plan extend to :
Also changing the DIAC nothing is working.
So this time the plan extend to :
- obtain the schematic (reverse engineering)
- remove again the transistors and verify they are still working, replace them with two tested MJE this time (now I have them)
- remove and change also the bridge (it implies to wait to receive a batch of new bridges)
- put the circuit in a test bench and VERY CAREFULLY, do some measurements
I want to see if at least the power arrives to the transistors and hopefully take also some measurements on the signals
All of this in order to identify other possible damaged components.
I also ordered some 6.8 uF 250V capacitors. They "seems" OK but at this point I have to assume the major critical components were fried !
I also ordered some 6.8 uF 250V capacitors. They "seems" OK but at this point I have to assume the major critical components were fried !
Round 3
Didn't have much time lately so I guess the circuit won :(
Did some measurements and the circuit seems doing something, I see some signals and the voltage seems OK, however the lamp remain OFF.
I had 3 new lamps and all 3 are not turning ON, probably one should try to play with different base resistors value.
For now I think I'll do other things, I did spend already enough time on that.
Testing
I'll put the new resistors not anymore on physical contact with the PCB.
Little bit more easier to dissipate the heat in the air rather than on the PCB for conduction.
Little bit more easier to dissipate the heat in the air rather than on the PCB for conduction.
Remember to cover the bulb with something because if it starts, it will emit dangerous UV !!
The tests are done with the container open.
Also be VERY careful to don't touch anything when powered since the circuit is connected to the power line directly ! No transformer.
The tests are done with the container open.
Also be VERY careful to don't touch anything when powered since the circuit is connected to the power line directly ! No transformer.
Of course in these cases, at least always power the circuit under test from a GFCI outlet !
So, after reconnecting the wires to the PCB time to see what happens !
Comment
Well, the reason I started the repair attempt is complicated.
Economically speaking to repair this kind of circuits is simply trashing away money and time.
Economically speaking to repair this kind of circuits is simply trashing away money and time.
But is quite sometime I don't play with electronics so when a friend of mine asked me to take a look, I decided to took the challenge, well knowing that very probably I would end up failing.
But ... is always fun to do so.
Too bad it didn't workout.
No comments:
Post a Comment